Kim Brace at Election Data Services has released his annual Congressional district projections based on the Census Bureau’s state population estimates for 2015:
More State’s Apportionment Allocations Impacted by New Census Estimates; New Twist in Supreme Court Case New Census Bureau population estimates for 2015 released today continue to show changes in states that will lead to adjustments in congressional apportionment.
The data shows eight states would already gain or lose additional congressional seats from what was officially assigned with the 2010 census, double that shown just last year in the 2014 estimates. In addition, this will grow to 15 states changing their congressional delegation size if the current trends continue to 2020. For the first time, the Bureau also released estimates of Voting Age Population and their use signals an interesting twist in the Supreme Court case argued earlier this month.
The Bureau’s 2015 total population estimates would add the states of Florida and Oregon to the list of states that will gain a seat, if the 2015 estimates were used for apportionment now. Illinois and Michigan join the list of states that are now likely to lose a congressional seat.
Last year’s 2014 estimates already determined that North Carolina and Texas would be gaining an additional district, while the states of Minnesota and Pennsylvania would lose a congressional district if apportionment was done with the new numbers. All other states would keep the same number of representatives they were awarded in December, 2010 when the official 2010 Census numbers were released.
While the Census Bureau has suffered budget cut-backs that have eliminated the production of state level population projections for upcoming decades, Election Data Services, Inc. has instead generated a simplified dataset by projecting forward the rates of change in populations from 2010 to 2015 reported by the Bureau within each state out to 2020.
The change in congressional delegations are the same if either the longer term trend of 2010 to 2015 is used, or a shorter trend of changes from 2013 to 2015 and 2014 to 2015 is utilized. Using this new set of projected 2020, the apportionment calculations show that 15 states could gain or lose districts by the time the Census is taken in 2020 in five years. The gainers and losers are:
States Gaining Districts
(6)Arizona +1 (from 9 to 10), Colorado +1 (from 7 to 8), Florida +2 (from 27 to 29), North Carolina +1 (from 13 to 14), Oregon +1 (from 5 to 6), Texas +3 (from 36 to 39).
States Loosing Districts (9)
Alabama -1 (from 7 to 6), Illinois -1 (from 18 to 17) , Michigan -1 (from 14 to 13), Minnesota -1 (from 8 to 7), New York -1 (from 27 to 26), Ohio -1 (from 16 to 15), Pennsylvania -1 (from 18 to 17), Rhode Island -1 (from 2 to 1), West Virginia -1 (from 3 to 2)
Last year’s population estimates indicated that both California and Virginia could have enough population to gain another seat in 2020, but the new Census Bureau data for 2015 and projected to 2020 shows those states just missing the cut.
California just missed gaining an additional seat in the new data, falling 29,302 people short at seat number 436 (there are just 435 congressional districts allocated to the states under a 1941 law capping the number of seats).
Virginia’s additional seat came in at seat number 437, missing the cut off by 69,841 people. The projections also demonstrate how close gaining states are to the magic 435 cut off.
Florida picked up their second district by capturing seat number 435 by only 15,608 people to spare.
The State of Arizona gained their congressional district at seat number 434 with only 13,741 people to spare.
Kimball Brace, President of Election Data Services, Inc. cautioned users to take the projections as very preliminary and subject to change, as evident by the California and Virginia new change. “We are only at the midpoint of the decade, and a lot of things could change before the next Census is taken in 2020,” Brace noted. “Having worked with Census data and estimates since the 1970s, it’s important to remember that major events like Katrina and the 2008 recession each changed population growth patterns and that impacted and changed the next apportionment,” he said.
Brace also noted that major changes in the counting process are being planned for 2020 and that reduced budget funding could impact those plans. “It would be ironic that Republican led efforts in the new Congress to cut government spending could cause Republican leaning states like Texas to lose out in apportionment,” said Brace. Texas is the big winner in the new projections, gaining as many as three districts in the study. But Brace also noted the irony in a new set of data released by the Census Bureau today.
For the first time the Bureau also released state estimates of voting age population (VAP) for 2015. The US Supreme Court earlier this month heard arguments in the Evenwel vs. Abbott court case where plaintiff’s argued that voters rather than total population should be used when drawing districts. Election Data Services ran the apportionment model using the VAP numbers for 2015 and discovered amongst other things that the State of Texas would actually lose a congressional district if voting age population were used for apportionment. These voting age numbers for 2015 track the gains and loses noted for other states in the above total population based apportionment study (gains of a seat in Florida, North Carolina and Oregon, against loses in states of Illinois, Minnesota and now Texas).
The Evenwel case challenges the Texas state legislative districts on one-person, one-vote grounds. The new 2015 estimates also point to how close a number of states stand to gain or lose a district. Most notable are the states of: Rhode Island – While keeping their two congressional districts with the 2015 numbers, the new data shows the state is now only 16,130 people away from dropping to a single district state. This is down from the 21,389 reported in last year’s apportionment study and from the 52,481 people margin they had in 2010.
At this rate they will be down to just one district in the next several years, the first time this has occurred to Rhode Island since 1789 when the nation was formed. They would join seven other states that also just have a single representative in the US House (Alaska, Delaware, Montana, North Dakota, South Dakota, Vermont and Wyoming) Oregon – The 2015 data indicates Oregon just gained its’ new congressional district.
The new estimates show the state gained seat number 435 by only 422 people to spare, one of the closest margins shown in our four decades of studying apportionment calculations. Michigan – On the flip side, the state of Michigan just lost their 14th congressional district, coming in at seat number 436 and missing the last seat to be handed out by just 1,038 people. Because congressional apportionment also impacts the Electoral College and the vote for President, Election Data Services took the 2020 projections for each state and applied the Presidential election results from the past four Presidential contests to determine the Electoral College outcomes in the past 15 years. The study shows that none of the presidential contests would have elected a different presidential candidate using the new apportionment counts but the would have been more Republican in nature. For example, in 2012 President Obama would still have won the Electoral College, but with three less votes (329 vs 332) that he won at the time of the voting. The biggest change would have occurred in the 2000 presidential election where George Bush would have gained an additional 19 electoral votes had the new 2020 apportionment projections determined the number of congressional seats in each state. The 2015 population estimates have not been statistically adjusted for any known undercount. No estimates were also not provided for U.S. military personnel overseas. This component has in the past been counted by the Census Bureau and allocated to the states. Overseas military personnel have been a factor in the apportionment formula for the past several decades, including the switching of the final district in 2000 that went from Utah to North Carolina. In both 1995 and 2005 the Census Bureau released population projections for states that went 25 years into the future. However, their website now says “The U.S. Census Bureau does not have a current set of state population projections and currently has no plans to produce them.” Earlier in 2014 the Bureau did release single nationwide population projections by age, sex, race and Hispanic origin for the time period of 2014 to 2060, but nothing by state. Past apportionment studies by Election Data Services, Inc. can be found at https://www.electiondataservices.com/reapportionment-studies/. A historical chart on the number of districts each state received each decade from 1789 to current is also available at this web address. Election Data Services Inc. is a political consulting firm that specializes in redistricting, election administration, and the analysis of census and political data. Election Data Services, Inc. conducts the congressional apportionment analyses with each annual release of the census population estimates.